On recovering non-local perturbation of non-selfadjoint Sturm-Liouville operator

TL;DR

This paper addresses the inverse spectral problem for a non-local, non-selfadjoint Sturm-Liouville operator with a frozen argument, developing conditions for unique potential recovery using spectrum and additional data.

Contribution

It extends inverse spectral theory to non-selfadjoint operators with non-local perturbations, providing necessary and sufficient spectral conditions and an algorithm for potential reconstruction.

Findings

Derived asymptotic spectral formulas for non-selfadjoint case

Identified uninformative parts of the spectrum for potential recovery

Established uniqueness and reconstruction algorithm using spectrum and additional data

Abstract

Recently, there appeared a significant interest in inverse spectral problems for non-local operators arising in numerous applications. In the present work, we consider the operator with frozen argument $ly = -y''(x) + p(x)y(x) + q(x)y(a),$ which is a non-local perturbation of the non-selfadjoint Sturm--Liouville operator. We study the inverse problem of recovering the potential $q\in L_2(0, \pi)$ by the spectrum when the coefficient $p\in L_2(0, \pi)$ is known. While the previous works were focused only on the case $p=0,$ here we investigate the more difficult non-selfadjoint case, which requires consideration of eigenvalues multiplicities. We develop an approach based on the relation between the characteristic function and the coefficients $\{ \xi_n\}_{n \ge 1}$ of the potential $q$ by a certain basis. We obtain necessary and sufficient conditions on the spectrum being asymptotic formulae of a special form. They yield that a part of the spectrum does not depend on $q,$ i.e. it is uninformative. For the unique solvability of the inverse problem, one should supplement the spectrum with a part of the coefficients $ \xi_n,$ being the minimal additional data. For the inverse problem by the spectrum and the additional data, we obtain a uniqueness theorem and an algorithm.